Divide your weekly conversions by 10. Run at least 4 and at most about 10 truly different creatives at once.
Published October 30, 2026 · By the Three Second Fame team
Divide the conversions you get in a week by 10. That is the number of creatives to run at once, with a minimum of 4 and a maximum of about 10.
Truly different means a different hook or a different angle. The same ad with a new colour counts as one creative.
Meta's guidance is about 50 conversions in 7 days per ad set, and the creatives in an ad set share those conversions. Spread that pool across 30 creatives and each one gets about 1.6 (50 ÷ 30 ≈ 1.6). One or two events cannot tell a good creative from a lucky one.
Producing 30 creatives is easy. Testing 30 at the same time produces noise.
Give a creative a verdict only after it has at least 10 target events. With only a few events, one extra or one missing event moves its cost per event a lot, so the number says little.
At 10 events per creative, the number of creatives you can judge is your weekly conversions divided by 10. That is where the formula comes from.
These are examples to show the arithmetic. Your weekly conversions will differ.
| Weekly conversions | Creatives to run | How it works out |
|---|---|---|
| 30 | 4 | The floor: 30 ÷ 10 is 3, and the minimum is 4. |
| 80 | 8 | 80 ÷ 10 is 8, between the floor and the ceiling. |
| 200 | About 10 | The ceiling: 200 ÷ 10 is 20, and the practical maximum is about 10. |
The ceiling exists because delivery is uneven, and the weakest creatives still need to reach the pause threshold. The floor exists because a test needs enough different ideas to map the space of possible angles.
Pause a creative that has spent 4 times the target cost per event and produced zero events.
Give every test at least 4 full days, however fast the events arrive. The minimum is there so a test covers days of the week that behave differently.
Never add the next batch before the current one is read. Adding creatives mid-test changes more than one thing at once, so you can no longer say which change caused a result. Pause the losers, then add the next batch to the same ad set.
Three Second Fame makes the creatives and puts the campaign up paused. Nothing is spent until you turn it on; only you can start spending. The budget behind these numbers is explained in how much to spend testing Meta ads. You can read how it works or see the pricing.
Creative counts here are arithmetic on the weekly conversions you get. The numbers in this article are working starting points from our method, not platform guarantees and not promises of results. Check them against your own data.